Annual Meeting - SCEC.org

Poster Abstracts
The data employed so far involves several aftershock sequences on small and known fault segments. These constrains allow
for better consideration of the path and site effects, including the path in the near vicinity of the source. Therefore, as opposed
to most models (using either the epicentral distance for low magnitude events or the closest surface projection of the fault for
stronger events) we consider the hypocentral source-receiver distance as controlling both the geometrical spreading and the
inelastic path effects. This approach is not only more physical for M < 5 events, but also reveals the upper crustal
heterogeneity (< 20km) in the San Jacinto Fault Zone area. The continuing work will focus on incorporating more data as it
becomes available, and using the results to develop updated Ground Motion Prediction Equations. The results will be useful
for earthquake engineering and for exploring the crustal heterogeneity around active faults.
LOCAL AND REGIONAL SEISMIC RESPONSE TO INJECTION AND PRODUCTION AT THE SALTON SEA
GEOTHERMAL FIELD, SOUTHERN CALIFORNIA (B-061)
L.J. Lajoie and E.E. Brodsky
California hosts both the largest geothermal resource capacity and highest seismicity rate in the nation. With plans to increase
geothermal output, and proven earthquake triggering in the vicinity of geothermal power plants worldwide, it is important to
determine the local and regional effects of geothermal power production. This study focuses on relating the volume of fluid
extracted from and re-injected into wells at the Salton Sea geothermal field (SSGF) in Southern California to local seismicity
rate and increased probability of larger events on nearby faults such as the San Andreas and Imperial faults. Seismic data is
obtained from the publicly available Advanced National Seismic System (ANSS) catalog and SSGF injection and production
data from the State of California Department of Conservation. We identify triggered earthquakes in the catalog by modeling
seismicity in a 15km radius around the SSGF according to an Epidemic-Type Aftershock Sequence (ETAS) method. The model
seeks to fit the cumulative seismicity curve from our dataset by optimizing five seismic parameters in accordance with
Gutenberg-Richter and Omori's law. The modeled curve is then removed from the dataset to isolate the non-ETAS, or
production-triggered, signal. We then formulate a constitutive law to relate the seismicity rate to the driving stress (i.e.
volumetric strain in the reservoir). Defining the local stressing rate provides a tool for predicting the effects that production
has on regional seismicity rates.
The largest spike in SSGF net production volume over the past 30 years is accompanied by the one of the largest increases in
both seismicity rate and moment release within the geothermal field. This indicates a direct coupling between net fluid
production volume (volume extracted minus volume re-injected) and seismicity rate and cumulative seismic moment in the
field. Three dimensional plots of hypocentral earthquake locations show that seismicity is concentrated on an approximately
NW-SE striking plane that dips shallowly to the west. Numerous inactive low angle normal faults with the same orientation
have been mapped within and bounding the Salton Trough, suggesting the fault underlying the SSGF could be a reactivated
detachment fault. Elevated seismicity rates on the detachment increases the probability of a larger earthquake on the fault that
could directly trigger an event on the San Andreas fault.
ELASTOSTATIC SOLUTIONS FOR REALISTIC SLIP AND STRESS AROUND MODE II AND III CRACKS (A-
111)
V.R. Lambert, S. Barbot, and J. Avouac
A common practice for analyzing the displacement and stress caused by slip on buried faults is to discretize the fault area into
finite size patches of assumed uniform slip and use the corresponding Green’s functions to predict strain at the surface or
within the bulk of the medium. This formalism is commonly used to invert observation of co-seismic deformation for fault slip,
in afterslip analyses, stress transfer studies and in numerical models of the seismic cycle, such as the Uniform California
Earthquake Rupture Forecast (UCERF). The discretization of the fault into areas of uniform slip introduces stress singularities
at the edges of each patch, which might make the results quite sensitive to the choice of the discretization scheme. Yet, little
attention has been paid to this issue so far. Here, we derive analytic expressions of the Green’s functions, for which the
Okada’s solution is a special case. We derive a full-space formulation for linearly interpolated fault slip distributions, which
allows discretization of any fault slip distributions without introducing stress singularities. We investigate the case of a finite
crack with a uniform distribution of stress on the fault. We show that the Okada solution provides an inadequate
representation of the stress distribution for mode II and III cracks, even in the limit of infinitely small discretization. We
introduce a critical size of fault discretization below which stress singularities dominate and bias the distribution. Our
tapered-slip solution does not suffer from these shortcomings, suppresses all singularities and converges to a uniform stress
with smaller discretization. Our results demonstrate the need to resolve the areas of stress concentration in our models of fault
slip evolution.
190 | Southern California Earthquake Center